294 F. M. BURNET 



all that is needed is a systematization of the results of recombination experi- 

 ments to indicate the extent to which characters can be modified or recom- 

 bined and the influence of experimental conditions on the results. 



(b) Genetic information must obviously be considered in any attempt to 

 provide a better picture of the processes taking place in the virus-infected 

 cell. It may be of value to those interested in the broader problem to attempt 

 to make clear some of the more immediate implications of the genetic findings. 



1. Genetic Aspects 



At the genetic level the main findings have already been outlined. Points 

 which may be stressed are as follows: 



(a) The existence of heterozygotes indicated that single infective particles 

 may contain two or more sets of genetic determinants (genomes). 



(b) Phenotypic mixtures of serotypes is a regular phenomenon, presumably 

 indicating that in the doubly infected cell there is no dominance of one set of 

 genomes over another. 



(c) Recombination is limited to interchange of two groups of characters; 

 hence no more than two genetic loci can be postulated. On the other hand, 

 characters within a single group can readily undergo mutational change 

 independently. 



(d) In our hands, heterozygotes are rapidly replaced by homozygotes. 



(e) The behavior of virulence in genetic interactions is complex. In some 

 examples, crosses between virulent and avirulent forms result in the appear- 

 ance of clones with a wide range of intermediate virulence. In other systems, 

 and particularly where a virulence character is linked to serological type, 

 no such redistribution of virulence occurs. 



Before discussing these points it is advisable to make some comment on 

 the symbols used. We believe that the most convenient convention is to 

 retain our present nomenclature for the phenotypic differences initially 

 between MEL and WSE using Roman letters; and to apply these if necessary 

 with distinguishing superscripts to phenotypic characters used in work with 

 other strains. Where discussion of genotype is required we believe that 

 until an accepted basis of theory is available the same symbols should be 

 used in italic script. In the case of MEL, the genotype will be ABDF-CE 

 without prejudice to future decision as to whether ABDF represents one 

 genetic locus or many. It is believed that this set of symbols could be modified 

 to deal with new facts and ideas as they developed. 



a. Heterozygosis. If one takes account of the findings in regard to hetero- 

 zygosis in both Hirst's laboratory and our own, certain likely conclusions can 

 be reached. Our results with the MEL/WSE system show that when fluids 

 are initiated with single infective units from the primary harvest of a 

 recombination experiment with a high level of phenotypic mixture, the great 



